ghjuvas/abrs

Program code for aspect-based review summarization

Aspect-based review summarization

Program code for aspect-based review summarization

About

This program performs aspect-based sentiment analysis using fine-tuned XLM-RoBERTa and summarization using clustering algorithm Affinity Propagation.

Technical

• Python3
• ABSA: torch, transformers
• Summarization: scikit-learn, scipy
• Normalization: torch, transformers

Requirements

You need Python3 environment. To install needed libraries, write in terminal from current folder this command:

pip3 install -r requirements.txt

Models

To perform ABSA you need model from this link:

XLM-RoBERTa

To perform normalization you need model from this link:

ruT5

Sample

To run code write following lines in terminal:

• python3 summarization.py corpus.txt -- to perform summarization
• python3 summarization.py -n corpus.txt -- to add normalization before printing

About me

Yanina Khudina, currently enrolled in bachelor program "Fundamential and Computational Linguistics" at HSE University, Moscow.

• e-mail: [email protected]